Cellular Cations Control Conformational Switching of Inositol Pyrophosphate Analogues

Article abstract of DOI:10.1002/chem.201601754

The inositol pyrophosphate messengers (PP-InsPs) are emerging as an important class of cellular regulators. These molecules have been linked to numerous biological processes, including insulin secretion and cancer cell migration, but how they trigger such a wide range of cellular responses has remained unanswered in many cases. Here, we show that the PP-InsPs exhibit complex speciation behaviour and propose that a unique conformational switching mechanism could contribute to their multifunctional effects. We synthesised non-hydrolysable bisphosphonate analogues and crystallised the analogues in complex with mammalian PPIP5K2 kinase. Subsequently, the bisphosphonate analogues were used to investigate the protonation sequence, metal-coordination properties, and conformation in solution. Remarkably, the presence of potassium and magnesium ions enabled the analogues to adopt two different conformations near physiological pH. Understanding how the intrinsic chemical properties of the PP-InsPs can contribute to their complex signalling outputs will be essential to elucidate their regulatory functions.

Full text of DOI:10.1002/chem.201601754

DOI: 10.1002/chem.201601754
Full Paper
&
Conformation Analysis |Hot Paper|
Cellular Cations Control Conformational Switching of Inositol
Pyrophosphate Analogues
Anastasia Hager,[b] Mingxuan Wu,[b] Huanchen Wang,[c] Nathaniel W. Brown, Jr.,[a, b]
Stephen B. Shears,[c] Nicolꢀs Veiga,*[d] and Dorothea Fiedler*[a, b]
Abstract: The inositol pyrophosphate messengers (PP-InsPs)
are emerging as an important class of cellular regulators.
These molecules have been linked to numerous biological
processes, including insulin secretion and cancer cell migra-
tion, but how they trigger such a wide range of cellular re-
sponses has remained unanswered in many cases. Here, we
show that the PP-InsPs exhibit complex speciation behaviour
and propose that a unique conformational switching mecha-
nism could contribute to their multifunctional effects. We
synthesised non-hydrolysable bisphosphonate analogues
and crystallised the analogues in complex with mammalian
PPIP5K2 kinase. Subsequently, the bisphosphonate ana-
logues were used to investigate the protonation sequence,
metal-coordination properties, and conformation in solution.
Remarkably, the presence of potassium and magnesium ions
enabled the analogues to adopt two different conformations
near physiological pH. Understanding how the intrinsic
chemical properties of the PP-InsPs can contribute to their
complex signalling outputs will be essential to elucidate
their regulatory functions.
Introduction
InsP5) by inositol hexakisphosphate kinases IP6K1, 2, and 3
(Figure 1a).[4] Subsequently, the ATP-grasp enzymes PPIP5K1
and PPIP5K2 further phosphorylate 5PP-InsP5 to yield 1,5-bis-
(diphosphoinositol) tetrakisphosphate (1,5(PP)2-InsP4, also
called InsP8, Figure 1a).[5] Compared with the InsP6 precursor,
the cellular concentrations of the inositol pyrophosphates 5PP-
Secondary messengers are key regulators within cellular signal
transduction networks. Related to the lipid-anchored phospha-
tidyl inositol signalling molecules, but less well characterised,
are the soluble inositol polyphosphates and inositol pyrophos-
phates (PP-InsPs).[1] The latter group of densely phosphorylated
molecules is particularly notable because the members bear
one or two highly energetic diphosphate groups attached to
the myo-inositol scaffold.[2] The large amount of energy neces-
sary for their biosynthesis, combined with their rapid cellular
turnover, strongly suggests a central signalling role for the PP-
InsPs.[3]
The canonical biosynthetic pathway for PP-InsPs in mamma-
lian cells depends on the action of two unrelated families of
small molecule kinases. First, inositol hexakisphosphate (InsP6)
is converted into 5-diphosphoinostiol pentakisphosphate (5PP-
[a] N. W. Brown, Jr., Prof. D. Fiedler
Leibniz-Institut fꢀr Molekulare Pharmakologie
Robert-Rçssle Strasse 10, 13125 Berlin (Germany)
E-mail: fiedler@fmp-berlin.de
[b] Dr. A. Hager, Dr. M. Wu, N. W. Brown, Jr., Prof. D. Fiedler
Department of Chemistry, Princeton University
Washington Rd., Princeton, New Jersey, 08544 (USA)
[c] Dr. H. Wang, Prof. S. B. Shears
Inositol Signaling Group, National Institutes of Health
Research Triangle Park, North Carolina, 27709 (USA)
[d] Prof. N. Veiga
Cꢁtedra de Quꢂmica Inorgꢁnica, Departamento Estrella Campos, Facultad
de Quꢂmica, Universidad de la Repfflblica, CC 1157, Montevideo (Uruguay)
E-mail: nveiga@fq.edu.uy
Figure 1. Inositol pyrophosphates and their corresponding bisphosphonate
analogues. a) Abbreviated biosynthetic pathway to generate the messengers
5-diphosphoinositol pentakisphosphate (5PP-InsP5), and 1,5-bis(diphospho-
inositol) tetrakisphosphate (1,5(PP)2-InsP4). b) Bisphosphonate analogues re-
ported here, and their applications.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601754.
Chem. Eur. J. 2016, 22, 1 – 10
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
InsP5 and 1,5(PP)2-InsP4 are low (InsP6: 30–50 mm, 5PP-InsP5: 1–
2 mm, 1,5(PP)2-InsP4: <1 mm), which may be attributed to the
rapid enzymatic hydrolysis of the PP-InsPs by diphosphoinosi-
tol polyphosphate phosphohydrolases (DIPPs).[2,6]
Characterisation of cell lines and organisms with perturbed
PP-InsP biosynthesis has uncovered a regulatory function of
these molecules in a wide range of cellular processes. For ex-
ample, in S. cerevisiae the PP-InsPs play a role in ribosome bio-
genesis,[7] telomere maintenance,[8] and phosphate sensing.[9] In
higher eukaryotes, the PP-InsPs are critical components in
cancer cell migration and metastasis,[10] insulin secretion,[11] in-
sulin sensitivity,[12] and host-cell immune response during viral
invasion.[13] How these messengers are able to elicit such
a broad spectrum of cellular responses has remained a topic of
active debate. Interestingly, the PP-InsPs can utilise two distinct
signalling mechanisms: protein binding and protein pyrophos-
phorylation.[2,7,14] In addition, the PP-InsPs are likely to interact
strongly with metal cations intracellularly,[15] which would
result in the formation of different PP-InsP-metal complexes. Fi-
nally, the highly phosphorylated inositols and 5PP-InsP5 were
shown to undergo a conformational change at high pH, which
could also contribute to their multifaceted outputs.[16]
Scheme 1. Synthesis of 5PCP-InsP5 (1).
devise a more efficient approach, utilising acetal 3 (Scheme 1).
This intermediate was reported by Holmes and co-workers,
and can be obtained from myo-inositol in two steps.[20] Along
this strategy, allylation of 3, followed by acetal cleavage,
formed allyl ether 4 in 84% yield. para-Methoxybenzyl (PMB)
protection of the hydroxyl groups and removal of the allyl pro-
tecting group provided reliable and scalable synthetic access
to alcohol 5, which was then converted into the bisphospho-
nate-containing intermediate 6. Despite the steric hindrance
around the secondary alcohol in 5, the bisphosphonate group
could be attached in good yield, using 4-(N,N-dimethylamino)-
pyridine (DMAP) and 5-phenyltetrazole (5-Ph-tetrazole) in di-
chloromethane. Removal of PMB groups from 6 furnished pen-
taol 7, which is an intermediate that we had previously em-
ployed in the synthesis of 5PCP-InsP5. Phosphitylation, oxida-
tion, and hydrogenolysis ultimately afforded 1 in good yield
and high purity.
This complex picture has sparked an interest in chemical
tools that can be utilised to unravel the underlying signalling
properties of the PP-InsPs. To this end, our group has pursued
the synthesis and characterisation of non-hydrolysable bi-
sphosphonate analogues of the PP-InsPs, and we previously
described bisphosphonate analogues 5PCP-InsP5 and 1PCP-
InsP5.[16e,17] Now we report improved synthetic access to 5PCP-
InsP5 (1), and the synthesis of bisphosphonate analogues for
both enantiomers of InsP8, 1,5(PCP)2-InsP4 and 3,5(PCP)2-InsP4
(Figure 1b, 2 and ent-2). The analogues were characterised
structurally by co-crystallisation with PPIP5K2 kinase, revealing
close structural mimicry of the bisphosphonate analogues
when compared with the natural molecules. Given that the bi-
sphosphonate group can withstand a wide pH range and is
not hydrolysed by the presence of Lewis acidic metal cati-
ons,[16e,17–19] we subsequently utilised the analogues to investi-
gate their protonation states, their cation chelating properties,
and their conformation in solution.
Synthesis of 1,5-(PCP)2-InsP4 and 3,5-(PCP)2-InsP4
Notably, potassium and magnesium ions had a significant in-
fluence on the conformational preferences of the bisphospho-
nate analogues, enabling the molecules to adopt two different
conformations near physiological pH. Our findings suggest
that a conformational switching mechanism could effectuate
multifunctional properties of the inositol pyrophosphate mes-
sengers in a cellular setting.
The synthesis of the bisphosphonate analogues of the InsP8
enantiomers 1,5-(PCP)2-InsP4 and 3,5-(PCP)2-InsP4 was of higher
synthetic complexity (Scheme 2). In the first step, the hydroxyl
group in acetal 3 was protected as tert-butyldimethylsilyl (TBS)
ether. Acetal cleavage delivered symmetric diol 8 in 68% yield
over two steps. The envisioned mono-PMB-protection of this
diol was more difficult than initially expected, and many reac-
tion conditions led to product mixtures or TBS deprotection.
Eventually, using phosphazene BEMP (2-tert-butylimino-2-dieth-
Results and Discussion
ylamino-1,3-dimethylperhydro-1,2,3-diazaphosphorine)
as
Synthesis of 5PCP-InsP5
a base, and freshly prepared PMB bromide, alcohol rac-9 could
be accessed in 65% yield. Removal of the TBS group provided
the unsymmetrical diol rac-10, as a key intermediate of the
synthesis, in almost quantitative yield. The subsequent attach-
ment of the bisphosphonate groups was challenging because
the diol system is highly sterically encumbered. In addition,
Our group previously reported the synthesis of bisphospho-
nate analogue 5PCP-InsP5 (1), in which the pyrophosphate
group is replaced by a bisphosphonate moiety.[16e] The synthe-
sis, although effective, involved an elaborate separation of re-
gioisomers early on in the procedure. We therefore sought to
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
plex with PPIP5K2KD, as well as the unnatural enantiomer
3,5(PP)2-InsP4, were reported recently.[21]
The structure of 5PCP-InsP5 (1) revealed that the analogue
binds to the highly positively charged active site (Figure 2a).
Superimposition of the structure with the corresponding cata-
lytic domain of PPIP5K2 in complex with 5PP-InsP5 showed
that the bisphosphonate analogue was captured in a very simi-
lar orientation, and the bisphosphonate group occupied the
same pocket as the diphosphate group of 5PP-InsP5 (Fig-
ure 2b). The highest deviation between the structures was
seen for the phosphate group in the 1-position. When the
carbon atoms of the inositol rings in the two structures were
superimposed, very good structural overlap was observed, and
the bisphosphonate group adopts a conformation almost iden-
tical to that of the diphosphate moiety (Figure 2c, RMSD=
0.041 ꢂ).
The structure of InsP8 analogue ent-2 in complex with
PPIP5K2 served to determine the absolute configuration of the
molecule. Enantiomer ent-2 turned out to be 3,5(PCP)2-InsP4;
the electron density map provided an unequivocal assign-
Scheme 2. Synthesis of 1,5-(PCP)2-InsP4 and 3,5-(PCP)2-InsP4.
the desired product features two additional stereogenic cen-
tres, resulting in a mixture of eight stereoisomers, which were
complex to purify. We ultimately developed a protocol that
relied on the continuous addition of an excess of the reagents,
allowing us to isolate bisphosphonate rac-11 in 28% yield as
a diastereomeric mixture. Global removal of PMB protecting
groups provided the tetraol intermediate, which could be
phosphitylated and oxidised in a one-pot procedure. The pro-
tected rac-(PCP)2-InsP4 was then hydrogenated, and subse-
quent ion-exchange provided access to the desired non-hydro-
lysable InsP8 mimic rac-2. To obtain the corresponding enantio-
mers, the racemic mixture of rac-10 was resolved by using
preparative chiral supercritical fluid chromatography (SFC).
Compound 10 and ent-10 were subsequently subjected to the
synthetic sequence described above, to yield enantiomerically
pure non-hydrolyzable analogues 2 and ent-2. Although the
optical rotations for 2 and ent-2 (see the Supporting Informa-
tion) were of similar magnitude but with opposite rotation, as-
signment of the absolute configuration solely based on com-
parison of these values to related compounds in the literature
is difficult. Given that optical rotation values can strongly
depend on pH and counter ion composition, we employed X-
ray crystallography to determine the absolute configuration.
Figure 2. Structural analysis of bisphosphonate analogues in complex with
PPIP5K2. a) Electrostatic surface presentation of the active site of PPIP5K2
with 5PCP-InsP5 bound. b) Superimposition of the catalytic domains of
PPIP5K2 bound to 5PCP-InsP5 (cyan CꢀC bonds) or 5PP-InsP5 (grey CꢀC
bonds and white spheres for all atoms). c) Superimposition of the carbon
atoms of the inositol ring in 5PCP-InsP5 (cyan CꢀC bonds) or 5PP-InsP5 (grey
CꢀC bonds and white spheres for all atoms). d) Electrostatic surface presen-
tation of the active site of PPIP5K2 with 3,5(PCP)2-InsP4 bound. e) Superim-
position of the catalytic domains of PPIP5K2 bound to 3,5(PCP)2-InsP4
(yellow CꢀC bonds) or 3,5(PP)2-InsP4 (white CꢀC bonds and white spheres
for all atoms). f) Superimposition of the carbon atoms of the inositol ring in
3,5(PCP)2-InsP4 (yellow CꢀC bonds) or 3,5(PP)2-InsP4 (white CꢀC bonds and
white spheres for all atoms).
Crystallographic analysis of PCP analogues bound to
PPIP5K2
To obtain structural information on the bisphosphonate ana-
logues, both 1 and ent-2 were co-crystallised with the kinase
domain of human PPIP5K2 (PPIP5K2KD). PPIP5K2 is a member
of an evolutionarily-conserved family of kinases responsible for
the conversion of 5PP-InsP5 into 1,5-(PP)2-InsP4,[5] and crystallo-
graphic data for both substrate and enzymatic product in com-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
ment.[22] Consequently, 2 must be 1,5-(PCP)2-InsP4, which was
subsequently confirmed by crystallisation (Figure S1). Just as in
the case of the natural molecules, the bisphosphonate ana-
logues were also accommodated in the positively charged
active site of the kinase (illustrated for 3,5(PCP)2-InsP4 in Fig-
ure 2d). Superimposition of kinase domain bound to either
3,5(PCP)2-InsP4 or 3,5(PP)2-InsP4 showed excellent overlap of
the analogue with 3,5(PP)2-InsP4 (Figure 2e). The structural
mimicry of the bisphosphonate analogue is further substantiat-
ed by the superimposition of the carbon atoms of the inositol
ring (Figure 2 f, RMSD=0.034 ꢂ), which shows that both bi-
sphosphonate groups closely align with the diphosphate
groups. The methylene group provides sufficient flexibility so
that the oxygen atoms of the b-phosphoryl moieties are well
superimposable with the oxygen atoms of the diphosphate
groups.
P3), 2 (P2), and 4 and 6 (P4/P6), as well as two signals corre-
sponding to the bisphosphonate moiety (P5a and P5b). The
changes in 31P NMR chemical shift with altered pH are depicted
in Figure 3a and 3b. With the exception of P5b, the signals
move downfield with increasing pH; an effect that was also
observed for the related molecule InsP6.[24]
Analysis of the NMR titration curves using HypNMR soft-
ware[25] provided the first seven protonation constants of fully
deprotonated 1 (L13ꢀ, Table S2). The solid lines in Figure 3a and
3b indicate the calculated chemical shifts of P1/P3, P2, P4/P6,
P5a and P5b, based on the derived protonation constants, and
illustrate an excellent fit to the experimental values. The corre-
sponding species distribution diagram (Figure 3c) shows that
the highly negatively charged H4L9ꢀ is the most abundant spe-
cies in solution near pH 7.2.[26]
Overall, the structural analysis of compounds 1 and ent-2 il-
lustrated a close resemblance of the analogues to the corre-
sponding diphosphate-containing molecules. Whereas the sub-
stitution of an oxygen atom for a methylene group can be as-
sociated with significant changes to the properties of a bi-
sphosphonate analogue, such as for example altered pKa
values and bond angles in the ATP analogue AMPPCP,[18,23] the
PCP-InsP analogues described here constitute very good
mimics from a structural perspective. The data thus bodes well
for future applications of the PCP-InsPs in crystallisation experi-
ments, or for the development of these analogues into affinity
reagents to identify protein binding partners.
pH-Dependent behaviour of 5PCP-InsP5 in solution in a non-
interacting medium
With good quantities of 5PCP-InsP5 (1) in hand, and having as-
certained that the compound constituted a good structural
mimic, we wanted to obtain a detailed view of the protonation
equilibria, the metal-coordination properties, and the confor-
mation of 1. A qualitative comparison of analogue 1 with natu-
rally occurring 5PP-InsP5 previously conducted by our group il-
lustrated that the behaviour of the two molecules in solution
is highly similar.[16e] PCP-analogue 1 is well suited for quantita-
tive thermodynamic studies, because a constant concentration
of 5PCP-InsP5 is required throughout the experiments, and the
bisphosphonate group is stable over a wide pH range, and is
resistant to hydrolysis mediated by Lewis acidic metal cati-
ons.[16e,17–19] Given the number of metal cations that have to be
considered in a biological setting, we decided to successively
introduce complexity into the system by investigating the
properties of 1 under the following conditions: a) in aqueous
solution, without coordinating metal cations present; b) with
added potassium ions; c) in the presence of potassium and
magnesium ions; d) under cellular concentrations of sodium,
potassium and magnesium.
To elucidate the protonation sequence of 1 in the absence
of coordinating counter ions, a 31P NMR titration was conduct-
ed at constant ionic strength (1 mm 1, 150 mm NMe4Cl,
pH 3.0–13.0). Due to the symmetry of 1, three resonances were
detected for the phosphate groups at positions 1 and 3 (P1/
Figure 3. NMR titration of 5PCP-InsP5 (1) without coordinating counter ions.
a) and b) 31P NMR chemical shifts of 1 as a function of pH, measured in
0.15m NMe4Cl at 22.08C. Solid lines indicate the chemical shifts predicted by
HypNMR, according to the protonation constants given in Table S1. c) Corre-
sponding species distribution diagram of 1. Predictions are for [L]=1.0 mm.
The pH range of the ligand conformational change is highlighted in grey.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The calculated 31P NMR chemical shifts for the different spe-
cies can be used to determine the protonation sites.[24] Given
that protonation has a shielding effect on the 31P nuclei, analy-
sis of the change in chemical shifts (DdP, Table S3) of P1/P3, P2,
P4/P6, P5a and P5b, during successive protonation provides
important information on the actual location of the protons.
For example, during the transition from HL12ꢀ to H2L11ꢀ, the
most negative DdP values were observed for P5a and P1/P3
(Table S3), indicating that these two phosphoryl groups were
protonated during this proton rearrangement step. The pre-
dicted most abundant microspecies and their protonation sites
are shown in Figure S3, along with a detailed explanation of
the procedure.
Notably, during the first protonation, the chemical shifts of
almost all resonances decreased significantly, suggesting a con-
formational change of the ligand. In its fully deprotonated
state, 5PCP-InsP5 was previously shown to adopt a conforma-
tion with five phosphate groups in axial positions and the 2-
phosphate group in an equatorial position (5a1e).[16e] Addition
of the first proton causes a flip from a 5a1e to a 1a5e (1 axial,
5 equatorial) conformation (Figure 4a), consistent also with the
broadening of the 31P NMR signals that is observed between
pH 11.5 and 12.5. The pH zone in which HL12ꢀ and L13ꢀ, and
thus both conformers, co-exist, is highlighted in the species
distribution diagram (Figure 3c). Interestingly, compared with
InsP6, bisphosphonate analogue 1 appears more recalcitrant to
undergo a conformational change to the 5a1e arrangement at
elevated pH.[24b]
Figure 4. pH-Dependent conformational change of 5PCP-InsP5 (1). a) Sche-
matic illustrating the transition from the 1a5e (1 axial, 5 equatorial) to 5a1e
(5 axial, 1 equatorial) conformation at elevated pH. b)–d) DFT optimised geo-
metries for the most stable conformers of L13ꢀ (b), HL12ꢀ (c), and H2L11ꢀ (d).
The intramolecular hydrogen bonds are shown as black dotted lines. Phos-
phoryl groups forming intramolecular hydrogen bonds are in bold. Colour
code: C (grey), H (white), O (red), P (orange).
Theoretical investigation of 5PCP-InsP5 protonation and con-
formation
To understand the conformational preferences of 1 in more
detail, the geometries of the predicted protonated species
were optimised and their energies were calculated by using
a RB3LYP/3-21+G* method (Figures 4 and S4). The fully depro-
tonated species L13ꢀ tends to adopt the 5a1e conformation,
because in this arrangement the highly negatively charged
phosphate groups are separated best (high XPꢀP values,
Table S4).[27] As discussed in the previous section, protonation
of L13ꢀ triggers a conformational change, so that the HL12ꢀ spe-
cies adopts the 1a5e form. The 1a5e conformation is stabilised
by strong intramolecular hydrogen bonds, which cause only
a small distortion of the inositol ring (Table S4). The calculated
structures suggest that two intramolecular hydrogen bonds
are important for stabilising the 1a5e conformation of HL12ꢀ
(Figure 4c); one hydrogen bond can form between P1 and P2
(1.22 ꢂ), another one between P4 and P5b (1.33 ꢂ). Compari-
son of the theoretical data to that obtained for InsP6 further il-
lustrates that the presence of the b-phosphoryl group has a sig-
nificant effect on both the protonation sequence and confor-
mation of the molecule. For InsP6, the transition from the 5a1e
to 1a5e conformation occurred during the HL11ꢀ to H2L10ꢀ pro-
tonation.[24b] This transition was observed experimentally, and
theoretical analysis of the energy difference between the two
conformers was fully consistent with the experimental results
(Figure S5). Analogous calculations for 1 revealed that the mol-
ecule only adopts the 5a1e conformation in its fully deproton-
ated state. Addition of the first proton triggers the conforma-
tional change, and the resulting HL12ꢀ species is stabilised
through intramolecular hydrogen bonds, including a unique
hydrogen bond between P4 and P5b (Figure 4c).
Coordination of potassium ions by 5PCP-InsP5 strongly influ-
ences inositol conformation
The characterisation of the speciation of 1 in a non-interacting
environment provided important structural information, such
as the stabilisation of the 1a5e conformation through intramo-
lecular hydrogen bonds. In a cellular environment, however,
numerous counter ions for the negatively charged inositol py-
rophosphates exist, and the presence of these counter ions
will have an impact on their structures and properties. We
therefore repeated the 31P NMR titrations in the presence of
150 mm KCl (a concentration close to physiological potassium
levels) over a pH range of 3.0 to 13.0. The titration curves im-
mediately indicated a significant interaction between 1 and
the potassium ions: Most 31P NMR signals were shifted down-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
field, due to a deshielding effect upon potassium complexation
(Figure S6). Mathematical analysis of the titration curves using
HypNMR provided an excellent fit to the experimental data
(Figure S6). Five K+ complexes were detected and their species
distribution is shown in Figure 5a. All the potassium com-
formation constant for [K4(H3L)]6ꢀ is higher than expected
when compared with the other equilibrium constants for suc-
cessive protonation (Table S2). The greater stability of
[K4(H3L)]6ꢀ is likely caused by the accompanying conformation-
al transition of the ligand.
The geometries of the most abundant potassium complexes
were then optimised by using DFT (Figure S8), and their ener-
gies were calculated. According to the computational data, the
5a1e conformation provides
a structure with negatively
charged cavities, which can be filled with potassium cations,
thereby alleviating the repulsion between phosphoryl groups
with only minor ring distortion. Only when the pH is low
enough will intramolecular H-bonds form to bring the phos-
phate groups into closer proximity and stabilise the 1a5e con-
formation. Furthermore, calculation of the energy difference
between the two conformations for the [K4(H2L)]7ꢀ versus the
[K4(H3L)]6ꢀ species fully supports the experimentally observed
conformational flip (Figure S8 f).
Overall, the strong metal-coordination properties of
1 caused a stark decrease in the pH range for the conforma-
tional change of the ligand, in this case provoked by a physio-
logical concentration of potassium ions (compare Figures 3c
and 5a).
Magnesium ions further stabilise the 5a1e conformation
Magnesium is the most abundant divalent cation in cells and it
has a propensity for forming strong coordination complexes
with phosphate-containing biomolecules, such as ATP and
InsP6.[28] We repeated the 31P NMR titration, now in a solution
containing 1 mm 1, 150 mm KCl, and 1 mm MgCl2, to provide
a cellular concentration of free magnesium cations, and to con-
duct the titration at a 1:1 stoichiometry of inositol pyrophos-
phate analogue to magnesium ion (Figure S9). Analysis of
these titration curves with HypNMR enabled us to determine
the formation constants associated with the magnesium/potas-
sium complexes of 1 (Table S2). The resulting chemical model
provided a good fit to the experimental data (Table S3, Fig-
ure S9 and S10) and is comprised of six highly stable 1:1 5PCP-
InsP5:Mg complexes, containing different numbers of protons
and potassium ions: [MgK4(HL)]6ꢀ, [MgK3(H2L)]6ꢀ, [MgK3(H3L)]5ꢀ,
[MgK2(H4L)]5ꢀ, [MgK(H5L)]5ꢀ and [Mg(H6L)]5ꢀ (Figure 5b). The lo-
cations of the protons and metal cations in these complexes
were determined from the DdP and DdC values from Table S3
(for detailed procedure see the Supporting Information). Based
on the solution thermodynamic data, the DFT optimised geo-
metries for four of these complexes are depicted in Figure 6.
The structures are stabilised mainly by a net of hydrogen and
coordinative bonds, which counteracts the strong electrostatic
repulsion between the negatively charged phosphate and bis-
phosphonate groups.
Figure 5. Species distribution diagrams of 5PCP-InsP5 (1) in the presence of
potassium and magnesium ions. a) [1]=1.0 mm, [K+]=150 mm.
b) [1]=1.0 mm, [K+]=150 mm, [Mg2+]=1.0 mm. T=22.08C. The pH range
of the ligand conformational change is highlighted in grey.
plexes were remarkably stable (Table S2), considering the low
ionic potential of potassium ions. Evidently, the high negative
charge of deprotonated 1 must contribute to the stability of
the potassium complexes.
The coordination and protonation sequence for the K+-con-
taining species can be elucidated by examining the calculated
individual chemical shifts and their variation with protonation
(DdP) and/or complexation (DdC, Table S3, and the Supporting
Information). Whereas protonation has a shielding effect, com-
plexation of potassium ions tends to shift the 31P resonances
downfield.[24b] The protonation and complexation processes
therefore become distinguishable. A comprehensive picture of
the predicted microspecies, and a description of the assign-
ments is provided in the Supporting Information (Figure S7). A
pronounced transition occurs when [K4(H2L)]7ꢀ is protonated to
yield [K4(H3L)]6ꢀ: all the 31P NMR signals are shifted significantly
(Table S3), revealing a pH-dependent conformational change of
the ligand. This conformational change is further confirmed ex-
perimentally by the broadening of the 31P NMR signals be-
tween pH 9.0 and 10.0, a pH range in which complexes
[K4(H2L)]7ꢀ and [K4(H3L)]6ꢀ co-exist (Figure 5a). Moreover, the
The inositol conformational change is operative between
pH 8.5 and 9.5, where the [MgK3(H3L)]5ꢀ complex is deproton-
ated to form the [MgK3(H2L)]6ꢀ species (Figure 5b). To confirm
that PCP analogue 1 indeed captures the behaviour of natural
5PP-InsP5 in solution accurately under the experimental condi-
tions described in this section, the 31P NMR spectra of 5PP-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
solic alkaline pH,[30,31] the species abundances change signifi-
cantly: the predominant species is [MgK3(H3L)]5ꢀ, with 49.7%
(Table S5), which is separated by a single deprotonation step
from the other conformer. In fact, a noticeable amount of
5PCP-InsP5 at pH 8.0 adopts the 5a1e state, as the
[MgK3(H2L)]6ꢀ complex (6.5%, Table S5). Therefore, provided
the appropriate cellular microenvironment, a fraction of 5PCP-
InsP5 could exist in the 5a1e conformation.
We next calculated the energy differences between the two
conformations for [MgK3(H3L)]5ꢀ and [MgK3(H2L)]6ꢀ, which were
experimentally found to occupy the 1a5e and 5a1e conforma-
tions, respectively (Figure 6b,c). Consistent with the NMR titra-
tion data, the [MgK3(H2L)]6ꢀ complex was most stable in the
5a1e state, whereas the 1a5e conformation constituted the
species of lowest energy for the [MgK3(H3L)]5ꢀ complex (Fig-
ure S12). Additional calculations indicated that the energy re-
quired for [MgK3(H3L)]5ꢀ to undergo the ring flip to the pre-
dominantly axial conformer is less than 10 kcalmolꢀ1. To place
that value in context, a comprehensive analysis of pharmaceut-
ically relevant protein–ligand complexes demonstrated that
large conformational rearrangements of the ligands (>9 kcal
molꢀ1) can be tolerated without penalising the tightness of
binding.[32] Therefore, it appears feasible that the energetic
cost for the conformational rearrangement could be compen-
sated for by the selective recognition of the 5a1e PP-InsP-Mg
complex by a protein binding partner. The [MgK3(H3L)]5ꢀ spe-
cies, which is quite abundant at physiological pH, may there-
fore target a unique set of protein binding partners in the
5a1e conformation, given that these proteins contain a binding
pocket that is complementary in charge and shape.
Figure 6. DFT optimised geometries for the detected 5PCP-InsP5 species in
complex with potassium and magnesium ions. Intramolecular hydrogen
bonds are shown as black dotted lines. Two unique hydrogen bonds from
P5b to a water molecule in the first coordination sphere of magnesium are
highlighted as green dotted lines. Phosphoryl groups forming intramolecular
hydrogen bonds are in bold. Colour code: C (grey), H (white), O (red), P
(orange), K (violet), Mg (lime green).
InsP5 were recorded between pH 6.0–12.0. Consistent with our
previous results,[16e] the conformational change in the presence
of potassium and magnesium ions occurs within the same pH
range for the bisphosphonate analogue and the natural mole-
cule, as evidenced by the chemical shifts for P2, P1/3, and P4/6
(Figure S11). Notably, the conformational change of 1 is shifted
by about 0.5 pH units towards physiological pH by the pres-
ence of one equivalent of magnesium ions. The stabilisation of
the 5a1e conformation can, in part, be attributed to the pre-
formed chelating site for magnesium ion between P4 and P6
(Figure 6a, b).
Conformational equilibrium of InsP6, 5PCP-InsP5, and rac-
(PCP)2-InsP4 in the presence of cellular metal ion concentra-
tions
Finally, we evaluated the conformational preferences of InsP6,
1, and rac-2 in a solution that contained both sodium and po-
tassium ions (10 mm NaCl, 130 mm KCl), and inositol polyphos-
phates at a lower concentration (125 mm, compared with 1 mm
in the previous experiments). To be able to detect the inositol
Theoretical investigations support this coordination scheme,
in which Mg2+ occupies a binding site between P4 and P6 in
the [MgK4(HL)]6ꢀ and [MgK3(H2L)]6ꢀ complexes. In addition, the
bisphosphonate group plays a stabilising role in the 5a1e
structures, by forming a hydrogen bond to one of the water
molecules in the first coordination sphere of the magnesium
ion (Figure 6a, b; green dotted lines). By comparison, the mag-
nesium coordination scheme is different in InsP6,[24b] again illus-
trating the distinctive features of the added b-phosphoryl
group.
1
species at these low concentrations, H NMR spectra were re-
corded, necessitating the use of deuterated solvent. The solu-
tions of InsP6, 1, and rac-2 were titrated from pD 5.0 to 11.5
(Figure S13). All molecules, including rac-2, underwent a confor-
mational change between pD 8.3 and 10.0, as illustrated by
the broadening of the signals in this range. We believe that, in
addition to the conformational change, chemical exchange be-
tween different metal-bound species further contributes to the
line broadening over this wide pH range.
At pH 7.2, the average pH in the cytoplasm and the nu-
One equivalent of MgCl2 (125 mm) was then added to the
samples, which shifted the range for the conformational
change towards physiological pH in all cases (Figure S13,
Figure 7). Given that InsP6 is known to form a complex of
higher stoichiometry with Mg2+ (1:5 InsP6:Mg2+),[28b] we won-
dered whether further addition of Mg2+ would affect the con-
formational equilibrium of InsP6 and the bisphosphonate ana-
logues. Indeed, addition of 2 or 3 equivalents of Mg2+ to the
cleus,[29] and in the presence of 150 mmK+ and 1 mmMg2+
,
the most abundant species are [MgK2(H4L)]5ꢀ (1a5e),
[MgK3(H3L)]5ꢀ (1a5e) and [K3(H4L)]6ꢀ (1a5e), at 50.8, 16.9 and
24.0%, respectively (Table S5). All of these species require one
or two deprotonation steps to undergo the conformational
change to the 5a1e state. When the pH is raised to 8.0, a pH
that can be encountered in spatially restricted regions of cyto-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
7
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
magnesium ions provoked a large shift in the conformational
equilibrium and facilitated the transition to the 5a1e conforma-
tion at much lower pH. Here, the metal cations coordinate to
the preformed cavities between the axial phosphate groups,
thereby minimising the electrostatic repulsion in the 5a1e con-
formation. In addition, theoretical investigations support
a unique magnesium binding site in 5PCP-InsP5, in which P4
and P6 directly coordinate to the metal, and P5b binds to
a water ligand in the first coordination sphere of the magnesi-
um ion, thereby further stabilising the 5a1e conformation. The
speciation diagram of the 5PCP-InsP5-Mg complexes revealed
that a small fraction of the inositol pyrophosphate analogue
can adopt a 5a1e conformation at pH 8.0, which is a pH that is
encountered in alkaline microdomains in the cytoplasm and in
the mitochondria. Importantly, theoretical analysis of the spe-
cies that are more abundant at physiological pH, such as
[MgK3(H3L)]5ꢀ, suggested that the energy required for promot-
ing a ring flip appears surmountable (<10 kcalmolꢀ1). There-
fore, a model in which specific interactions of the PP-InsP-
metal complexes with target proteins induce the stabilisation
of the 5a1e conformation appears plausible, and some of
these interactions may be metal-mediated. The fascinating
question then arises whether the PP-InsPs can target different
sets of cellular proteins, depending on their metal-complexa-
tion and their conformation.
Figure 7. Effect of magnesium ions on the conformational equilibrium of
InsP6, 5PCP-InsP5 (1) and rac-(PCP)2-InsP4 (rac-2). Schematic illustrating a shift
of the conformational change of InsP6, 1 and rac-2, towards physiological
pD upon addition of magnesium.
InsP6 solutions triggered the conformational transition at lower
pD (Figure S14). For samples of 1 and rac-2, the addition of 2
or 3 equivalents of Mg2+ also resulted in broadening of the
1H NMR signals at lower pD, between 7.0 and 7.5. The elevated
amount of magnesium ions unfortunately caused precipitation
of the samples at higher pD, so that the 5a1e conformation for
1 and rac-2 never became visible under these conditions (Fig-
ure S15). Overall, the complexation of magnesium ions by the
highly phosphorylated inositols facilitated the conformational
transition at a lower pH, closer to the physiological environ-
ment. While it is tempting to speculate that the inositol pyro-
phosphates—when in solution at cellular concentrations (1–
2 mm), and when surrounded by cellular levels of free magnesi-
um ions (0.5–1.5 mm)—are able to access both conformations
with high frequency at physiological pH, this hypothesis can
only be tested with more sensitive spectroscopic measure-
ments. These measurements could, for example, rely on incor-
porating 13C into the inositol ring,[33] or could employ more ad-
vanced techniques such as dynamic nuclear polarization
(DNP)[34] to lower the limit of detection.
In sum, a picture emerges in which the PP-InsPs can achieve
their multifunctional outputs by a variety of mechanisms. Cer-
tainly, the spatially and temporally controlled production of
these messengers by specific kinases can bestow different sig-
nalling properties onto these molecules. Intriguingly though,
the PP-InsPs could also control distinct cellular signalling
events in response to changes in their protonation, metal com-
plexation, and conformation in the cell. The latter features are
not dependent on the action of specific biosynthetic enzymes,
but are instead dictated by the intrinsic chemical properties of
the PP-InsPs, and the environment and/or the protein binding
partners they encounter. Elucidating these added layers of
complexity will be essential to understand the regulatory func-
tions of inositol pyrophosphate messengers within cellular sig-
nalling networks.
Conclusion
We have synthesised and characterised bisphosphonate ana-
logues of the cellular messengers 5PP-InsP5 and 1,5-(PP)2-InsP4.
The characterisation included crystallographic studies, which
demonstrated high structural overlap between the non-hydro-
lysable analogues and the diphosphate containing molecules.
The methylene group of the bisphosphonate moiety possesses
a high degree of flexibility, which enables it to adopt a confor-
mation very similar to that of the diphosphate group.
Experimental Section
Experimental details, including full characterization for all com-
pounds, as well as supporting tables and figures, are provided in
the Supporting Information.
Analysis of the protonation constants, metal-coordination
properties, and conformation of the bisphosphonate ana- Acknowledgements
logues illustrated their complex behaviour in solution, and
comparison to the properties of InsP6 highlighted the distinc-
tive features of the b-phosphoryl group: In the absence of co-
ordinating counter ions, 5PCP-InsP5 was stabilised in a 1a5e
conformation by the formation of strong intramolecular hydro-
gen bonds, in particular a hydrogen bond between P5b and
P4/P6. Introduction of cellular concentrations of potassium and
We thank Dr. Istvan Pelczer, Ken Conover, and Dr. Peter
Schmieder for guidance with the NMR experiments, and their
helpful suggestions. We also thank Laura Wilson from Lotus
Separations for assistance with the chiral SFC. Anastasia Hager
gratefully acknowledges the Deutsche Forschungsgemein-
schaft (DFG) for her postdoctoral fellowship. D.F. received fund-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
8
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
A. Saiardi, Proc. Natl. Acad. Sci. USA 2009, 106, 21161–21166; c) Z. Szij-
gyarto, A. Garedew, C. Azevedo, A. Saiardi, Science 2011, 334, 802–805.
[15] In analogy to other inositol polyphosphate species: a) N. Veiga, J.
Torres, S. Domꢃnguez, A. Mederos, R. F. Irvine, A. Dꢃaz, C. Kremer, J.
Inorg. Biochem. 2006, 100, 1800–1810; b) J. Torres, N. Veiga, J. S. Gan-
cheff, S. Domꢃnguez, A. Mederos, M. Sundberg, A. Sꢀnchez, J. Castiglio-
ni, A. Dꢃaz, C. Kremer, J. Mol. Struct. 2008, 874, 77–88; c) N. Veiga, J.
Torres, H. Y. Godage, A. M. Riley, S. Domꢃnguez, B. V. L. Potter, A. Dꢃaz, C.
Kremer, J. Biol. Inorg. Chem. 2009, 14, 1001–1013.
[16] a) L. R. Isbrandt, R. P. Oertel, J. Am. Chem. Soc. 1980, 102, 3144; b) J. C.
Lindon, D. J. Baker, R. D. Farrant, J. M. Williams, Biochem. J. 1986, 233,
275; c) C. Blum-Held, P. Bernard, B. Spiess, J. Am. Chem. Soc. 2001, 123,
3399; d) A. M. Riley, M. Trusselle, P. Kuad, M. Borkovec, J. Cho, J. H. Choi,
X. Qian, S. B. Shears, B. Spiess, B. V. L. Potter, ChemBioChem 2006, 7,
1114; e) M. Wu, B. E. Dul, A. J. Trevisan, D. Fiedler, Chem. Sci. 2013, 4,
405–410.
ing from the NIH (DP2 CA186753), the Sidney Kimmel Founda-
tion, the Rita Allen Foundation, and Princeton University.
Keywords: biological activity
·
conformation analysis
·
phosphorylation · protonation · signal transduction
[1] a) R. F. Irvine, M. J. Schell, Nat. Rev. Mol. Cell. Biol. 2001, 2, 327–338;
b) R. F. Irvine, J. Physiol. 2005, 566, 295–300.
[2] a) S. B. Shears, Adv. Biol. Regul. 2015, 57, 203–216; b) M. S. Wilson, T. M.
Livermore, A. Saiardi, Biochem. J. 2013, 452, 369–379; c) T. Wunden-
berg, G. W. Mayr, Biol. Chem. 2012, 393, 979–998; d) A. Chakraborty, S.
Kim, S. H. Snyder, Science Signal. 2011, 4, re1–11; e) C. J. Barker, C. Illies,
G. C. Gaboardi, P.-O. Berggren, Cell. Mol. Life Sci. 2009, 66, 3851–3871;
f) J. P. Monserrate, J. D. York, Curr. Opin. Cell Biol. 2010, 22, 365–373.
[3] a) L. Stephens, T. Radenberg, U. Thiel, G. Vogel, K. H. Khoo, A. Dell, T. R.
Jackson, P. T. Hawkins, G. W. Mayr, J. Biol. Chem. 1993, 268, 4009–4015;
b) F. S. Menniti, R. N. Miller, J. W. Putney, S. B. Shears, J. Biol. Chem. 1993,
268, 3850–3856; c) M. C. Glennon, S. B. Shears, Biochem. J. 1993, 293,
583–590.
[17] M. Wu, L. S. Chong, S. Capolicchio, H. J. Jessen, A. C. Resnick, D. Fiedler,
Angew. Chem. Int. Ed. 2014, 53, 9508–9511; Angew. Chem. 2014, 126,
9662–9665.
[18] T. S. Elliott, A. Slowey, Y. Ye, S. J. Conway, Med. Chem. Commun. 2012, 3,
735–751.
[4] a) A. Saiardi, H. Erdjument-Bromage, A. M. Snowman, P. Tempst, S. H.
Snyder, Curr. Biol. 1999, 9, 1323–1326; b) A. Saiardi, E. Nagata, H. R.
Luo, A. M. Snowman, S. H. Snyder, J. Biol. Chem. 2001, 276, 39179–
39185; c) M. J. Schell, A. J. Letcher, C. A. Brearley, J. Biber, H. Murer, R. F.
Irvine, FEBS Lett. 1999, 461, 169–172; d) S. M. Voglmaier, M. E. Bembe-
nek, A. I. Kaplin, G. Dorman, J. D. Olszewski, G. D. Prestwich, S. H.
Snyder, Proc. Natl. Acad. Sci. USA 1996, 93, 4305–4310.
[5] a) S. Mulugu, W. Bai, P. C. Fridy, R. J. Bastidas, J. C. Otto, D. E. Dollins,
T. A. Haystead, A. A. Ribeiro, J. D. York, Science 2007, 316, 106–109;
b) J. H. Choi, J. Williams, J. Cho, J. R. Falck, S. B. Shears, J. Biol. Chem.
2007, 282, 30763–30775; c) H. Lin, P. C. Fridy, A. A. Ribeiro, J. H. Choi,
D. K. Barma, G. Vogel, J. R. Falck, S. B. Shears, J. D. York, G. W. Mayr, J.
Biol. Chem. 2008, 284, 1863–1872.
[19] L. M. Yates, D. Fiedler, ChemBioChem 2015, 16, 415–423.
[20] S. J. Conway, J. Gardiner, S. J. Grove, M. K. Johns, Z. Y. Lim, G. F. Painter,
D. E. Robinson, C. Schieber, J. W. Thuring, L. S. Wong, M. X. Yin, A. W.
Burgess, B. Catimel, P. T. Hawkins, N. T. Ktistakis, L. R. Stephens, A. B.
Holmes, Org. Biomol. Chem. 2010, 8, 66–76.
[21] a) H. Wang, J. R. Falck, T. M. Hall, S. B. Shears, Nat. Chem. Biol. 2011, 8,
111–116; b) S. Capolicchio, H. Wang, D. T. Thakor, S. B. Shears, H. J.
Jessen, Angew. Chem. Int. Ed. 2014, 53, 9508–9511; Angew. Chem. 2014,
126, 9662–9665.
[22] Position 5 in 3,5-(PCP)2-InsP4 shows alternative conformations in the
solid state (Figure S1).
[23] G. M. Blackburn, D. E. Kent, F. Kolkmann, J. Chem. Soc. Chem. Commun.
1981, 1188–1190.
[6] a) J. J. Caffrey, S. T. Safrany, X. Yang, S. B. Shears, J. Biol. Chem. 2000, 275,
12730–12736; b) L. V. Hua, K. Hidaka, X. Pesesse, L. D. Barnes, S. B.
Shears, Biochem. J. 2003, 373, 81–89; c) R. S. Kilari, J. D. Weaver, S. B.
Shears, S. T. Safrany, FEBS Lett. 2013, 587, 3464–3470.
[7] a) C. Horigome, R. Ikeda, T. Okada, K. Takenami, K. Mizuta, Biochemistry
2009, 73, 443–446; b) S. G. Thota, C. P. Unnikannan, S. R. Thampatty, R.
Manorama, R. Bhandari, Biochem. J. 2015, 466, 105–114.
[24] a) A. Bebot-Brigaud, G. Dange, N. Fauconnier, C. Gꢄrard, J. Inorg. Bio-
chem. 1999, 75, 71–78; b) N. Veiga, J. Torres, I. Macho, K. Gomez, G.
Gonzalez, C. Kremer, Dalton Trans. 2014, 43, 16238–16251.
[25] G. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M. S. Moruzzi, A. Vacca, Anal.
Biochem. 1995, 231, 374–382.
[26] The titration was also conducted at 378C. The species distribution dia-
gram did not change significantly (Figure S2).
[8] a) S. J. York, B. N. Armbruster, P. Greenwell, T. D. Peters, J. D. York, J. Biol.
Chem. 2005, 280, 4264–4269; b) A. Saiardi, A. C. Res-nick, A. M. Snow-
man, B. Wendland, S. H. Snyder, Proc. Natl. Acad. Sci. USA 2005, 102,
1911–1914.
[9] Y. S. Lee, S. Mulugu, J. D. York, E. K. O’Shea, Science 2007, 316, 109–112.
[10] F. Rao, J. Xu, C. Fu, J. Y. Cha, M. M. Gadalla, R. Xu, J. C. Barrow, S. H.
Snyder, Proc. Natl. Acad. Sci. USA 2015, 112, 1773–1778.
[11] a) C. Illies, J. Gromada, R. Fiune, B. Leibiger, J. Yu, K. Juhl, S.-N. Yang,
D. K. Barma, J. R. Falck, A. Saiardi, C. J. Barker, P. O. Berggren, Science
2007, 318, 1299–1302; b) R. Bhandari, K. R. Juluri, A. C. Resnick, S. H.
Snyder, Proc. Natl. Acad. Sci. USA 2008, 105, 2349–2353.
[12] A. Chakraborty, M. A. Koldobskiy, N. T. Bello, M. Maxwell, J. J. Potter, K. R.
Juluri, D. Maag, S. Kim, A. S. Huang, M. J. Dailey, M. Saleh, A. M. Snow-
man, T. H. Moran, E. Mezey, S. H. Snyder, Cell 2010, 140, 897–898.
[13] N. K. Pulloor, S. Nair, K. McCaffrey, A. D. Kostic, P. Bist, J. D. Weaver, A. M.
Riley, R. Tyagi, P. D. Uchil, J. D. York, S. H. Snyder, A. Garcꢃa-Sastre, B. V.
Potter, R. Lin, S. B. Shears, R. J. Xavier, M. N. Krishnan, PLoS Pathog.
2014, 10, e1003981.
[27] The calculated energy for the corresponding 1a5e conformer is not
much higher, which could explain the persistent broadening of the
31P NMR resonances above pH 11.
[28] a) W. Jahnen-Dechent, M. Ketteler, Clin. Kidney J. 2012, 5, i3–i14; b) J.
Torres, S. Domꢃnguez, M. F. Cerdꢀ, G. Obal, A. Mederos, R. F. Irvine, A.
Dꢃaz, C. Kremer, Eur. J. Inorg. Chem. 2005, 99, 828–840.
[29] J. R. Casey, S. Grinstein, J. Orlowski, Nat. Rev. Mol. Cell Biol. 2010, 11, 50–
61.
[30] a) C. J. Schwiening, D. Willoughby, J. Physiol. 2002, 538, 371–382;
b) H. A. Ro, J. H. Carson, J. Biol. Chem. 2004, 279, 37115–37123.
[31] R. Bizzarri, C. Arcangeli, D. Arosio, F. Ricci, P. Faraci, F. Cardarelli, F. Bel-
tram, Biophys. J. 2006, 90, 3300–3314.
[32] E. Perola, P. S. Charifson, J. Med. Chem. 2004, 47, 2499–2510.
[33] A. Saiardi, C. Guillermier, O. Loss, J. C. Poczatek, C. Lechene, Surf. Inter-
face Anal. 2014, 46, 169–172.
[34] Q. Z. Ni, E. Daviso, T. V. Can, E. Markhasin, S. K. Jawla, T. M. Swager, R. J.
Temkin, J. Herzfeld, R. G. Griffin, Acc. Chem. Res. 2013, 46, 1933–1941.
[14] a) R. Bhandari, A. Saiardi, Y. Ahmadibeni, A. M. Snowman, A. C. Resnick,
T. Z. Kristiansen, H. Molina, A. Pandey, J. K. Werner Jr., K. R. Juluri, Y. Xu,
G. D. Prestwich, K. Parang, S. H. Snyder, Proc. Natl. Acad. Sci. USA 2007,
104, 15305–15310; b) C. Azevedo, A. Burton, E. Ruiz-Mateos, M. Marsh,
Received: April 14, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
9
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Conformation Analysis
A. Hager, M. Wu, H. Wang,
N. W. Brown, Jr., S. B. Shears, N. Veiga,*
D. Fiedler*
&& – &&
Mixed messages: The inositol pyro-
phosphate messengers (PP-InsPs) are
emerging as an important class of cellu-
lar regulators. Herein, it was shown that
the PP-InsPs exhibit complex speciation
behaviour and it is proposed that
a unique conformational switching
mechanism could contribute to their
multifunctional effects (see figure).
Cellular Cations Control
Conformational Switching of Inositol
Pyrophosphate Analogues
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
10
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!